Joint Energy Management and Eco-Routing for Electric Vehicles with Hybrid Energy Storage Systems

The eco-routing problem for battery/ultracapacitorbased electric vehicles is investigated in this paper to minimize the total energy cost, and a reinforcement learning-based joint energy management and eco-routing (JEMER) algorithm is proposed, which ensures that our eco-routing can cope with dynamic traffic in real time. Then the optimization and real-time performances of the energy management strategy (EMS) are proved by standard driving cycles, and the JEMER algorithm is validated on the Sioux Falls network. Our results show that the JEMER algorithm can provide an EMS that balances optimal control and real-time application, and also reveal that significant energy savings can be achieved with a small increase in travel time. Additionally, as the traffic flow increases, the extent of time sacrifice increases significantly while the energy saving decreases slightly; therefore, drivers should decide whether to choose the eco-route according to the road traffic flow and their time generosity.